Human Cognitive Reliability Model Research Articles

Handling Uncertainty in Human Cognitive Reliability Method for Safety Assessment Based on DSET

Human Reliability Analysis (HRA) is an important part in safety assessment of a large complex system. Human Cognitive Reliability (HCR) model is a method of evaluating the probability that operators fail to complete during diagnostic decision making within a limited time, which is widely used in HRA. In the application of this method, cognitive patterns of humans are required to be considered and classified, and this process often relies on the evaluation opinions of experts which is highly subjective and uncertain. How to effectively express and process this uncertain and subjective information plays a critical role in improving the accuracy and applicability of HCR. In this paper, a new model was proposed to deal with the uncertain information which exists in the processes of cognitive pattern classification in HCR. First, an evaluation panel was constructed based on expert opinions and processing including setting corresponding anchor points and qualitative indicators of different cognitive patterns, and mapping them to fuzzy numbers and unit intervals. Second, based on the evaluation panel, different analysts judge the cognitive pattern types of actual specific events and provide the level of confidence he or she has in the judgments. Finally, the evaluation opinions of multiple analysts were expressed and fused based on the Dempster-Shafer Evidence Theory (DSET), and the fused results were applied to the HCR model to obtain the Human Error Probability (HEP). A case study was used to demonstrate the procedure and effectiveness of the proposed method.

The simulator experimental study on the operator reliability of Qinshan nuclear power plant

The scarcity of human failure data is a major problem that embarrasses human reliability analysts for a long time. In the human reliability analysis of Qinshan nuclear power plant, a full-size simulator was used to conduct an experiment on the operator's reliability. The experiment involves 23 abnormal events and 55 human interactions covering 3 cognitive types: skill-based, rule-based and knowledge-based. The response time and conditions of 38 operators were recorded and the data were analyzed and processed to develop human cognitive reliability model parameters, which match with the features of Qinshan nuclear power plant system and its staff. This paper introduces the background, theory, methods, process, conclusion of the experiment and their comparison with data obtained abroad.

Cognitive model research of nuclear power plant operators

Human cognitive reliability (HCR) model is a very useful quantitative model to the reliability research of nuclear power plant (NPP) operators based on simulators. The event non-response probability versus normalized time curve of HCR can be evaluated by statistical distributions such as three-parameter Weibull distribution and lognormal distribution. The paper analyses the HCR model and the characteristics of Chinese NPP operators. Firstly, two-parameter Weibull distribution is used to evaluate the fitting curve of HCR. Then the reliability of NPP operators and the verification of HCR model are studied. The research shows not only the quantitative results of Chinese NPP operators but also a new method to verify HCR model, and can be referred to the real NPP operation, and hence, is of importance to safe operation of NPP.

Mathematical Description of Crew Response Times in Simulated Nuclear Power Plant Emergencies

Response time data were collected from experienced nuclear power plant crews in a pressurized water reactor simulator. Four different estimation techniques were used to fit three-parameter Weibull distributions to the data. Fits were excellent with the four techniques showing good agreement. Results yielded only weak support for the Human Cognitive Reliability model that relates these fitted distributions to skill-, knowledge-, and rule-based behaviors.

A Comparison of Operator Response Times Predicted by the HCR Model with those Obtained from Simulators

A case study of the analysis of operator reponse times following an abnormal event is presented. A number of timings of operator responses on a full‐scale nuclear plant simulator have been taken. These timings are compared with the human cognitive reliability (HCR) model. Agreement with the basic framework of the model is shown. Based on the data collected and on operators′ judgements on behaviour differences on simulators and the plant, criticisms of some aspects of the model are raised. The principal findings are: (1) the technique relies on the judgement of the analyst in the correct selection of behaviour type. The results of the analysis are very sensitive to this behaviour type; (2) the methods proposed for estimating the median response time are questioned, particularly when applying data from simulator measurements to the plant.

Assessing Crew Response Probability for a Sequence of Actions with an Overall Time Constraint

The following situation has been considered: during an accident sequence, a crew has to do a proper set of actions within a finite time window in order to achieve a specific goal. The operator actions are considered independent, they are lined up in series, and a single failure leads to mission failure. This paper proposes a general methodology for assessing the crew response (or success) probability, through a convolution integral formulation. This method is general and can be applied to any response probability model; an application to the Human Cognitive Reliability model is given.

Application of the human cognitive reliability model and confusion matrix approach in a probabilistic risk assesmeent

Abstract A general methodology for quantifying nonresponse human error rates, which makes use of the human cognitive reliability (HCR) model, has been developed. A step-by-step procedure is also described for the investigation of errors of misdiagnosis. The objective is to ensure traceability, repeatability and consistency in the level of analysis for each action. This methodology is suitable for application in a full-scope, probabilistic risk assessment (PRA).

A critique of recent models for human error rate assessment

Abstract This paper critically reviews two groups of models for assessing human error rates under accident conditions. The first group, which includes the US Nuclear Regulatory Commission (NRC) handbook model and the human cognitive reliability (HCR) model, considers as fundamental the time that is available to the operators to act. The second group, which is presented by the success likelihood index methodology-multiattribute utility decomposition (SLIM-MAUD) model, relies on ratings of the human actions with respect to certain qualitative factors and the subsequent derivation of error rates. These models are evaluated with respect to two criteria: the treatment of uncertainties and the internal coherence of the models. In other words, this evaluation focuses primarily on normative aspects of these models. The principal findings are as follows: (1) Both of the time-related models provide human error rates as a function of the available time for action and the prevailing conditions. However, the HCR model ignores the important issue of state-of-knowledge uncertainties, dealing exclusively with stochastic uncertainty, whereas the model presented in the NRC handbook handles both types of uncertainty. (2) SLIM-MAUD provides a highly structured approach for the derivation of human error rates under given conditions. However, the treatment of the weights and ratings in this model is internally inconsistent.

The role of frameworks, models, data, and judgment in human reliability analysis

Many advancements in the methods for treating human interactions in PRA studies have occurred in the last decade. These advancements appear to increase the capability of PRAs to extend beyond just the assessment of the human's importance to safety. However, variations in the application of these advanced models, data, and judgment in recent PRAs make quantitative comparisons among studies extremely difficult. This uncertainty in the analysis diminishes the usefulness of the PRA study for upgrading procedures, enhancing training, simulator design, technical specification guidance, and for aid in designing the man-machine interface. Hence, there is a need for a framework to guide analysts in incorporating human interactions into the PRA systems analyses so that future users of a PRA study will have a clear understanding of the approaches, models, data, and assumptions which were employed in the initial study. This paper describes the role of the systematic human action reliability procedure (SHARP) in providing a road map through the complex terrain of human reliability analysis that promises to improve the reproducibility of such analysis in the areas of selecting the models, data, representations, and assumptions. Also described is the role that a human cognitive reliability model can have in collecting data from simulators and helping analysts assign human reliability parameters in a PRA study. Use of these systematic approaches to perform or upgrade existing PRAs promises to make PRA studies more useful as risk management tools.